HPV testing compared with routine cytology in cervical screening: long-term follow-up of ARTISTIC RCT.

Abstract

The National Screening Committee (NSC) based its recommendation that human papillomavirus (HPV) testing should replace cytology in primary cervical screening largely on the 2009 follow-up results of the ARTISTIC trial (A Randomised Trial In Screening To Improve Cytology). The NSC must now decide on screening intervals and triage policy. Options include extending the screening interval up to 10 years for human papillomavirus-negative (HPV-) women, delaying recall for human papillomavirus-positive (HPV+) women with normal cytology (as their infections are usually transient), and basing triage on full HPV typing. In ARTISTIC, 24,510 women were recruited who were attending routine cervical cytology in Greater Manchester in 2001-3. The women were randomly allocated between revealing and concealing their HPV test results and were recalled every 3 years. After 2009, the women returned to routine cytological screening with recall every 3 years for those aged < 50 years, and every 5 years for those aged 50-64 years. We have followed the cohort to 2015 through national cancer registration for CIN3 (cervical intraepithelial neoplasia grade 3) and cancer, and through linkage to the cervical screening call-recall system to obtain lifetime cytology records. The analysis comprised 24,496 women at round 1 and 13,591 women at round 2 (which was 30-48 months later). Follow-up via local histology laboratories and national cancer registration identified 505 cases of cervical intraepithelial neoplasia grade 3 or cervical cancer (CIN3+) (including 22 invasive cervical cancers). The cumulative CIN3+ risk 10 years after a negative HPV test [0.31%, 95% confidence interval (CI) 0.18% to 0.49%, in the revealed arm] was similar to that 3 years after negative cytology (0.30%, 95% CI 0.23% to 0.41%, in the concealed arm) and fell sharply with age, from 1.1% (95% CI 0.7% to 1.8%) in those women aged < 25 years to 0.08% (95% CI 0.03% to 0.20%) in those women aged > 50 years. The 10-year cumulative CIN3+ risk following a new HPV infection at round 2 was 3.4% (95% CI 2.1% to 5.4%). The highest risks were associated with type-specific persistent infections that, overall, resulted in a 10-year cumulative CIN3+ risk of 20.4% (95% CI 15.6% to 26.4%). We found a similar level of protection 10 years after a negative HPV test and 3 years after negative cytology. These data support a considerably longer screening interval after a negative HPV test than after a negative cytology test. About three-quarters of women with HPV infection and normal cytology clear their infections within about 3 years. Their risk of CIN3+ within this time frame is low (1.5%), suggesting that the current policy of annual repeat testing and referral after 2 years may be unnecessarily cautious. Approximately 40% of women who remained HPV+ had cleared their initial infection and acquired a new HPV type. The cumulative CIN3+ risks in women with type-specific persistent infections are about six times higher than in women with new infections. Triage strategies based on HPV persistence would, therefore, reduce unnecessary referral of women with new (and largely transient) infections. HPV assays that identify HPV types 31, 33, 45, 52 and 58 in addition to 16 and 18 could be useful in triage as well as in primary HPV testing. Similar results in recent routine HPV screening suggest that our results are generalisable despite changes in cytology and HPV assay methods. We are continuing to follow the ARTISTIC cohort into the new era of primary HPV screening. Future work will focus on the implications of more sensitive HPV testing for primary HPV screening policy and triage of HPV-positive women. Our results suggest that a more sensitive test is needed to detect occult CIN3 at high risk of progression to cancer, but this would substantially increase the overall HPV detection rate. Tests such as DNA (deoxyribonucleic acid) methylation for distinguishing HPV infection from neoplasia will be evaluated on stored samples and on further samples now being collected from women in the cohort who are still being screened. This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 23, No. 28. See the NIHR Journals Library website for further project information.